submission_id: hastagaras-llama3-8b-darktune_v2
developer_uid: Hastagaras
status: inactive
model_repo: Hastagaras/llama3-8b-darktune
reward_repo: ChaiML/reward_gpt2_medium_preference_24m_e2
generation_params: {'temperature': 0.8, 'top_p': 0.9, 'min_p': 0.0, 'top_k': 40, 'presence_penalty': 0.0, 'frequency_penalty': 0.0, 'stopping_words': ['<|eot_id|>', '\n{bot_name}:', '\n{user_name}:'], 'max_input_tokens': 512, 'best_of': 16, 'max_output_tokens': 64}
formatter: {'memory_template': "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n\nAct as {bot_name}. {bot_name}'s persona: {memory}<|eot_id|>", 'prompt_template': '<|start_header_id|>Assistant<|end_header_id|>\n\n{bot_name}: {prompt}<|eot_id|>', 'bot_template': '<|start_header_id|>Assistant<|end_header_id|>\n\n{bot_name}: {message}<|eot_id|>', 'user_template': '<|start_header_id|>user<|end_header_id|>\n\n{user_name}: {message}<|eot_id|>', 'response_template': '<|start_header_id|>Assistant<|end_header_id|>\n\n{bot_name}:', 'truncate_by_message': False}
reward_formatter: {'memory_template': "{bot_name}'s Persona: {memory}\n\n", 'prompt_template': '{prompt}\n<START>\n', 'bot_template': '{bot_name}: {message}\n', 'user_template': '{user_name}: {message}\n', 'response_template': '{bot_name}:', 'truncate_by_message': False}
timestamp: 2024-05-09T09:48:41+00:00
model_name: hastagaras-lm3-8b-darktune_v1
model_eval_status: success
double_thumbs_up: 75
thumbs_up: 139
thumbs_down: 68
num_battles: 8981
num_wins: 4906
celo_rating: 1189.37
entertaining: 7.2
stay_in_character: 8.75
user_preference: 7.48
safety_score: 0.95
submission_type: basic
model_architecture: LlamaForCausalLM
model_num_parameters: 8030261248.0
best_of: 16
max_input_tokens: 512
max_output_tokens: 64
display_name: hastagaras-lm3-8b-darktune_v1
double_thumbs_up_ratio: 0.26595744680851063
feedback_count: 282
ineligible_reason: None
language_model: Hastagaras/llama3-8b-darktune
model_score: 7.81
model_size: 8B
reward_model: ChaiML/reward_gpt2_medium_preference_24m_e2
single_thumbs_up_ratio: 0.4929078014184397
thumbs_down_ratio: 0.24113475177304963
thumbs_up_ratio: 0.7588652482269503
us_pacific_date: 2024-05-09
win_ratio: 0.5462643358200646
Resubmit model
Running pipeline stage MKMLizer
Starting job with name hastagaras-llama3-8b-darktune-v2-mkmlizer
Waiting for job on hastagaras-llama3-8b-darktune-v2-mkmlizer to finish
hastagaras-llama3-8b-darktune-v2-mkmlizer: ╔═════════════════════════════════════════════════════════════════════╗
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ _____ __ __ ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ / _/ /_ ___ __/ / ___ ___ / / ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ / _/ / // / |/|/ / _ \/ -_) -_) / ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ /_//_/\_, /|__,__/_//_/\__/\__/_/ ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ /___/ ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ Version: 0.8.10 ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ Copyright 2023 MK ONE TECHNOLOGIES Inc. ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ The license key for the current software has been verified as ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ belonging to: ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ Chai Research Corp. ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ Account ID: 7997a29f-0ceb-4cc7-9adf-840c57b4ae6f ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ Expiration: 2024-07-15 23:59:59 ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ║ ║
hastagaras-llama3-8b-darktune-v2-mkmlizer: ╚═════════════════════════════════════════════════════════════════════╝
hastagaras-llama3-8b-darktune-v2-mkmlizer: /opt/conda/lib/python3.10/site-packages/huggingface_hub/utils/_deprecation.py:131: FutureWarning: 'list_files_info' (from 'huggingface_hub.hf_api') is deprecated and will be removed from version '0.23'. Use `list_repo_tree` and `get_paths_info` instead.
hastagaras-llama3-8b-darktune-v2-mkmlizer: warnings.warn(warning_message, FutureWarning)
hastagaras-llama3-8b-darktune-v2-mkmlizer: Downloaded to shared memory in 21.809s
hastagaras-llama3-8b-darktune-v2-mkmlizer: quantizing model to /dev/shm/model_cache
hastagaras-llama3-8b-darktune-v2-mkmlizer: Saving flywheel model at /dev/shm/model_cache
hastagaras-llama3-8b-darktune-v2-mkmlizer: Loading 0: 0%| | 0/291 [00:00<?, ?it/s] Loading 0: 42%|████▏ | 121/291 [00:01<00:01, 119.56it/s] Loading 0: 85%|████████▍ | 247/291 [00:02<00:00, 123.31it/s] Loading 0: 99%|█████████▊| 287/291 [00:07<00:00, 30.19it/s] Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
hastagaras-llama3-8b-darktune-v2-mkmlizer: quantized model in 17.396s
hastagaras-llama3-8b-darktune-v2-mkmlizer: Processed model Hastagaras/llama3-8b-darktune in 40.326s
hastagaras-llama3-8b-darktune-v2-mkmlizer: creating bucket guanaco-mkml-models
hastagaras-llama3-8b-darktune-v2-mkmlizer: Bucket 's3://guanaco-mkml-models/' created
hastagaras-llama3-8b-darktune-v2-mkmlizer: uploading /dev/shm/model_cache to s3://guanaco-mkml-models/hastagaras-llama3-8b-darktune-v2
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /dev/shm/model_cache/special_tokens_map.json s3://guanaco-mkml-models/hastagaras-llama3-8b-darktune-v2/special_tokens_map.json
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /dev/shm/model_cache/config.json s3://guanaco-mkml-models/hastagaras-llama3-8b-darktune-v2/config.json
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer_config.json s3://guanaco-mkml-models/hastagaras-llama3-8b-darktune-v2/tokenizer_config.json
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /dev/shm/model_cache/tokenizer.json s3://guanaco-mkml-models/hastagaras-llama3-8b-darktune-v2/tokenizer.json
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /dev/shm/model_cache/flywheel_model.0.safetensors s3://guanaco-mkml-models/hastagaras-llama3-8b-darktune-v2/flywheel_model.0.safetensors
hastagaras-llama3-8b-darktune-v2-mkmlizer: loading reward model from ChaiML/reward_gpt2_medium_preference_24m_e2
hastagaras-llama3-8b-darktune-v2-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py:913: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
hastagaras-llama3-8b-darktune-v2-mkmlizer: warnings.warn(
hastagaras-llama3-8b-darktune-v2-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py:757: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
hastagaras-llama3-8b-darktune-v2-mkmlizer: warnings.warn(
hastagaras-llama3-8b-darktune-v2-mkmlizer: /opt/conda/lib/python3.10/site-packages/transformers/models/auto/auto_factory.py:468: FutureWarning: The `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.
hastagaras-llama3-8b-darktune-v2-mkmlizer: warnings.warn(
hastagaras-llama3-8b-darktune-v2-mkmlizer: /opt/conda/lib/python3.10/site-packages/torch/_utils.py:831: UserWarning: TypedStorage is deprecated. It will be removed in the future and UntypedStorage will be the only storage class. This should only matter to you if you are using storages directly. To access UntypedStorage directly, use tensor.untyped_storage() instead of tensor.storage()
hastagaras-llama3-8b-darktune-v2-mkmlizer: return self.fget.__get__(instance, owner)()
hastagaras-llama3-8b-darktune-v2-mkmlizer: Saving model to /tmp/reward_cache/reward.tensors
hastagaras-llama3-8b-darktune-v2-mkmlizer: Saving duration: 0.257s
hastagaras-llama3-8b-darktune-v2-mkmlizer: Processed model ChaiML/reward_gpt2_medium_preference_24m_e2 in 3.081s
hastagaras-llama3-8b-darktune-v2-mkmlizer: creating bucket guanaco-reward-models
hastagaras-llama3-8b-darktune-v2-mkmlizer: Bucket 's3://guanaco-reward-models/' created
hastagaras-llama3-8b-darktune-v2-mkmlizer: uploading /tmp/reward_cache to s3://guanaco-reward-models/hastagaras-llama3-8b-darktune-v2_reward
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /tmp/reward_cache/config.json s3://guanaco-reward-models/hastagaras-llama3-8b-darktune-v2_reward/config.json
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /tmp/reward_cache/merges.txt s3://guanaco-reward-models/hastagaras-llama3-8b-darktune-v2_reward/merges.txt
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /tmp/reward_cache/special_tokens_map.json s3://guanaco-reward-models/hastagaras-llama3-8b-darktune-v2_reward/special_tokens_map.json
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /tmp/reward_cache/vocab.json s3://guanaco-reward-models/hastagaras-llama3-8b-darktune-v2_reward/vocab.json
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /tmp/reward_cache/tokenizer_config.json s3://guanaco-reward-models/hastagaras-llama3-8b-darktune-v2_reward/tokenizer_config.json
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /tmp/reward_cache/tokenizer.json s3://guanaco-reward-models/hastagaras-llama3-8b-darktune-v2_reward/tokenizer.json
hastagaras-llama3-8b-darktune-v2-mkmlizer: cp /tmp/reward_cache/reward.tensors s3://guanaco-reward-models/hastagaras-llama3-8b-darktune-v2_reward/reward.tensors
Job hastagaras-llama3-8b-darktune-v2-mkmlizer completed after 62.69s with status: succeeded
Stopping job with name hastagaras-llama3-8b-darktune-v2-mkmlizer
Pipeline stage MKMLizer completed in 65.35s
Running pipeline stage MKMLKubeTemplater
Pipeline stage MKMLKubeTemplater completed in 0.17s
Running pipeline stage ISVCDeployer
Creating inference service hastagaras-llama3-8b-darktune-v2
Waiting for inference service hastagaras-llama3-8b-darktune-v2 to be ready
Inference service hastagaras-llama3-8b-darktune-v2 ready after 40.204068183898926s
Pipeline stage ISVCDeployer completed in 47.17s
Running pipeline stage StressChecker
Received healthy response to inference request in 2.163341760635376s
Received healthy response to inference request in 1.3966999053955078s
Received healthy response to inference request in 1.307760238647461s
Received healthy response to inference request in 1.4027771949768066s
Received healthy response to inference request in 1.358346939086914s
5 requests
0 failed requests
5th percentile: 1.3178775787353516
10th percentile: 1.3279949188232423
20th percentile: 1.3482295989990234
30th percentile: 1.3660175323486328
40th percentile: 1.3813587188720704
50th percentile: 1.3966999053955078
60th percentile: 1.3991308212280273
70th percentile: 1.4015617370605469
80th percentile: 1.5548901081085207
90th percentile: 1.8591159343719483
95th percentile: 2.011228847503662
99th percentile: 2.132919178009033
mean time: 1.525785207748413
Pipeline stage StressChecker completed in 8.38s
Running pipeline stage DaemonicModelEvalScorer
Pipeline stage DaemonicModelEvalScorer completed in 0.03s
Running pipeline stage DaemonicSafetyScorer
Running M-Eval for topic stay_in_character
Pipeline stage DaemonicSafetyScorer completed in 0.03s
M-Eval Dataset for topic stay_in_character is loaded
hastagaras-llama3-8b-darktune_v2 status is now deployed due to DeploymentManager action
hastagaras-llama3-8b-darktune_v2 status is now inactive due to auto deactivation removed underperforming models

Usage Metrics

Latency Metrics